NCL hangs

From: <Oliver.Fuhrer_at_nyahnyahspammersnyahnyah>
Date: Mon, 27 Jul 2009 15:34:20 +0200

Hi all!

I'm trying to compute some power spectra of kinetic energy from model
output. I use NCL to compute the spectra and save the results in an
ASCII file. No graphics are done. I run the script over several hours of
model output and average the results. If I run the time loop only over
one or two hours (setting the variable hstop=6 or 7 in the script) the
script runs fine. If I let the loop run over the full forecast starting
at +6h up to +24h (by setting hstop=24) NCL simply hangs and doesn't
continue. Looking at "top" on the machine, NCL is using 100% CPU. This
should not be a memory issue, as according to "top" NCL is using only
~100 MB and the node has 8GB available. The 3 fields that are read in
from the Grib1 file have the dimensions 520x350x60.

Thanks for any help,
Oli

- I've attached the script to this email

- I am using the following NCL version...

5.1.1

- I am running on the following machine...

Linux buin2 2.6.5-7.283-ss #4 SMP Fri Sep 28 13:24:48 PDT 2007 x86_64
x86_64 x86_64 GNU/Linux
 
- The example (including code, data and grib tables) can be found
here...

https://twiki.cscs.ch/twiki/bin/viewfile/Main/OliverFuhrerDownloads?rev=
1;filename=ncl_bug.tar.gz

- Instructions to run...

> /bin/tcsh
> cd ncl_bug
> source setenv.csh
> ncl comp_spec.ncl

- The script gives the following output when running into the hang...

...
(0) debug 60
(0) debug 70
(0) debug 80
(0) debug 90
(0) Opening file data/lfff00230000
warning:NclGRIB: Unknown grib parameter ...
warning:NclGRIB: Unknown grib parameter ...
warning:NclGRIB: Unknown grib parameter ...
warning:NclGRIB: Unknown grib parameter ...
warning:NclGRIB: Unknown grib parameter ...
warning:NclGRIB: Unknown grib parameter ...
(0) working on direction E-W
(0) detrending and tapering
(0) computing FFT
(0) computing kinetic energy density
(0) debug 10
(0) debug 20
(0) debug 30
(0) debug 40
(0) debug 50
(0) debug 60
(0) debug 70
(0) debug 80
(0) working on direction N-S
(0) detrending and tapering
(0) computing FFT
(0) computing kinetic energy density
(0) debug 10
(0) debug 20
(0) debug 30
(0) debug 40
(0) debug 50
(0) debug 60
(0) debug 70
(0) debug 80
(0) debug 90

________________________________________

Oliver Fuhrer
Numerical Models

Federal Departement of Home Affairs FDHA
Federal Office of Meteorology and Climatology MeteoSwiss

Kraehbuehlstrasse 58, P.O. Box 514, CH-8044 Zurich, Switzerland

Tel. +41 44 256 93 59
Fax +41 44 256 92 78
oliver.fuhrer_at_meteoswiss.ch
www.meteoswiss.ch - First-hand information
  

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk

Received on Mon Jul 27 2009 - 07:34:20 MDT

This archive was generated by hypermail 2.2.0 : Wed Jul 29 2009 - 09:07:41 MDT